Google, Microsoft, Meta, Amazon, OpenAI, LinkedIn, Adobe, and dating app company Match Group (parent company of Tinder and Hinge) have jointly signed a new commitment called the "Online Services Accord Against Scams," which is expected to involve sharing threat intelligence and working together to combat fraudulent activities using their platforms. This agreement comes ahead of the UN Global Fraud Summit in Austria.Exclusively revealed by AxiosThis symbolizes a crucial step of unity in the tech industry as it faces the growing threat of fraud.
Why join forces now? AI has caused fraud losses to surge past $160 billion.
This agreement stems from a stark reality: fraud is no longer a lone wolf crime. According to the FBI, global consumers lost over $160 billion to fraud and cybercrime in 2024. The widespread adoption of generative AI allows fraud groups to easily overcome language barriers, generating highly personalized and virtually indistinguishable phishing emails, fake customer service conversations, and even deepfake videos, rendering traditional defense mechanisms ineffective.
In an interview with Axios, Karen Courington, VP of Consumer Trust Experience at Google's Trust and Security team, frankly stated, "We can't solve this problem alone. We need the entire industry to unite and address fraud in a more collective way." This statement highlights the core spirit of the agreement: facing organized and professional transnational fraud chains, blocking efforts from a single platform will only allow criminal groups to shift their operations; only cross-platform information sharing can build an effective firewall.
The agreement's four pillars are: intelligence sharing, AI defense, payment verification, and user protection.
According to a draft agreement obtained by Axios, the signatory companies pledged to take action in four main areas:
• Enhanced intelligence sharing mechanisms: Companies will share threat intelligence about transnational criminal networks more closely with industry peers and law enforcement agencies. This is not limited to investigations of individual cases, but also includes exchanging best practices for preventing and detecting fraud through international forums such as the Global Anti-Scam Alliance and the Tech Against Scams Coalition.
• Deploy AI-based defensive tools: A commitment to developing and deploying new defensive tools, particularly leveraging AI systems to detect fraud patterns more quickly. This means that future AI may not only be used for fraud but will also become a key weapon in the fight against fraud.
• Strengthen financial transaction verification: Implement a stricter verification mechanism for financial transactions on the platform to ensure the authenticity of both parties in the transaction and cut off the profit channel of fraud groups from the financial transaction end.
• Provide accessible reporting channels and promote legislation: Provide clear reporting channels for users who have been scammed, and call on governments to formally "make fraud prevention a national priority".
From "case-by-case collaboration" to "regular alliance": An upgrade in information sharing
Nathaniel Gleicher, Global Head of Security Policy and Anti-Fraud at Meta, pointed out that while companies had previously exchanged intelligence on individual cases, this agreement provides a platform for regular communication that goes beyond single cases. They can now discuss more systematically "which defenses are effective?" and "how are threat actors changing their tactics?"
Steven Masada, Assistant General Counsel of Microsoft's Digital Crime Division, emphasized that this partnership allows for more immediate communication between partners. He anticipates launching more disruptive operations in the future to "more effectively destroy infrastructure and identify threat actors harming customers worldwide."
Scott Knapp, Amazon's Vice President of Global Buyer Risk Prevention, pointed out the cross-platform nature of scams: "Scammers rarely operate on a single platform; they often perform the same operations across multiple online profiles." For example, a scammer might contact a victim on a social media platform, then direct them to another messaging app, and finally use a dating app to build trust. This cross-platform crime chain can only be completely dismantled through cross-platform cooperation.
Behind the Agreement: The Effectiveness of its Voluntary Nature and Regulatory Pressure
However, this agreement is not a legally binding treaty, but rather a voluntary commitment, and it doesn't even set up any penalties for companies that violate the agreement. This inevitably raises questions about its binding force.
However, from another perspective, the signing of this agreement itself carries strong symbolic significance. It sends a signal that tech companies are willing to elevate "anti-fraud" to the level of strategic cooperation, rather than merely engaging in public relations. Furthermore, with regulations such as the EU's Digital Services Act (DSA) imposing increasingly stringent requirements on platform responsibility, failure to take proactive measures could lead to harsher legal sanctions and hefty fines in the future. Therefore, this agreement can also be seen as a "preventative defense" by the tech industry under regulatory pressure.
It is worth noting that in the final week before the agreement was signed, Meta and OpenAI announced new anti-fraud measures, including the launch of new security features and the establishment of partnerships, indicating that major companies had already prepared for the agreement.
Analysis of viewpoints
The signing of this international agreement represents a fundamental shift in the mindset of the technology industry in combating fraud, moving the fight against fraud from an "arms race" to an "alliance era."
First, the shift from "passive prevention" to "proactive hunting." In the past, corporate anti-fraud measures were often reactive, meaning investigations were only conducted after users had been victimized. By sharing threat intelligence, companies now have the opportunity to identify potential threats from the infrastructure, domains, or payment patterns used by fraud groups before they launch attacks, enabling "proactive hunting." Microsoft's announced "more effective disruption operations" are a concrete manifestation of this strategic shift.
Secondly, the "asymmetry between offense and defense" in AI is about to be broken. Fraudulent groups use AI to launch large-scale attacks, but now defenders will also begin to deploy AI on a large scale for detection. This will usher in an "AI vs. AI" offensive and defensive battle. Future AI models will not only need to understand semantics, but also be able to identify subtle fraudulent intentions, emotional manipulation patterns, and even cross-platform fake identity associations.
Furthermore, "payment verification" will become the focus of the next wave of regulation and innovation. The agreement specifically mentions strict verification of financial transactions, which means that in the future, whether sending gifts on dating apps, purchasing high-priced goods on e-commerce platforms, or clicking on advertisements on social media, users may face more cumbersome identity verification processes. How to strike a balance between "security" and "user experience" will be a challenge that all signatory companies must face.
Finally, this agreement also highlights the harsh reality that "cross-platform crime" requires "cross-platform alliances" to solve. A fraudulent account may be deleted on Facebook, but the perpetrator could be looking for their next target on Tinder. With the addition of Match Group, this agreement successfully connects social media and dating apps—the two most common venues for fraud rings—completing a crucial piece of the puzzle.
Overall, the signing of the Online Services Anti-Fraud Agreement is not only a moral statement by tech giants, but also a strategic alliance to address threats of the new era. While the details and implementation remain to be seen, it does paint a vision of a safer internet for billions of users worldwide.
In the future, we will see more security features centered on "fraud prevention" launched, as well as more active lobbying by major companies to government regulatory agencies, working together to corner global fraud groups.
